Polar Biol (2008) 31:1461–1468 DOI 10.1007/s00300-008-0487-z
ORIGINAL PAPER
Satellite tracking reveals distinct movement patterns for Type B and Type C killer whales in the southern Ross Sea, Antarctica Russel D. Andrews · Robert L. Pitman · Lisa T. Ballance
Received: 27 March 2008 / Revised: 17 June 2008 / Accepted: 17 June 2008 / Published online: 30 July 2008 © Springer-Verlag 2008
Abstract During January/February 2006, we satellitetracked two diVerent ecotypes of killer whales (Orcinus orca) in McMurdo Sound, Ross Sea, Antarctica, using surface-mounted tags attached with sub-dermal darts. A single Type B whale (pinniped prey specialist), tracked for 27 days, traveled an average net distance of 56.8 § 32.8 km day¡1, a maximum of 114 km day¡1, and covered an estimated area of 49,351 km2. It spent several days near two large emperor penguin (Aptenodytes forsteri) colonies, a potential prey item for this form. By contrast, four Type C killer whales (Wsh prey specialists) tracked for 7–65 days, traveled an average net distance of 20 § 8.3 km day¡1, a maximum of 56 net km day¡1, and covered an estimated area of only 5,223 km2. These movement patterns are consistent with those of killer whale ecotypes in the eastern North PaciWc where mammal-eating ‘transients’ travel widely and are less predictable in their movements, and Wsh-eating ‘residents’ have a more localized distribution and more predictable occurrence, at least during the summer months. Keywords Antarctica · Killer whale · Ecotype · Ross Sea · Satellite tracking
R. D. Andrews (&) University of Alaska Fairbanks School of Fisheries and Ocean Sciences, and the Alaska SeaLife Center, P.O. Box 1329, Seward, AK 99664, USA e-mail:
[email protected] R. L. Pitman · L. T. Ballance NOAA Fisheries, Southwest Fisheries Science Center, 8604 La Jolla Shores Dr., La Jolla, CA 92037, USA
Introduction Long-term research on killer whales (Orcinus orca) in the Northeast PaciWc and Antarctica has revealed that they can comprise populations of up to three sympatric, largely noninterbreeding ecotypes that specialize on diVerent prey types (Ford et al. 1998; Baird 2000; Pitman and Ensor 2003). Furthermore, speciWc prey preferences among these ecotypes appear to be linked to diVerent foraging habitats and movement patterns (Ford et al. 2000). In the eastern North PaciWc, for example, the relationship between feeding and movement patterns has led to three ecotypes being referred to as ‘transients’, a marine mammal-eating form; ‘residents’, a Wsh-eating form that specializes on salmonids, and ‘oVshores’, a less well-known form that spends little time in coastal areas and whose food habits are largely unknown but do include Wsh (Bigg 1982; Ford et al. 2000; Baird 2000; Jones 2006). All three North PaciWc killer whale ecotypes have overlapping geographic ranges (at least during summer), but residents occupy smaller ranges and are more predictably located than either transients or oVshores. Recently, three diVerent killer whale ecotypes were described from Antarctica based on Weld observations of color pattern diVerences, apparent prey specialization, habitat and herd size diVerences (Pitman and Ensor 2003), as well as diVerences in body length for at least two of the forms (Pitman et al. 2007). The three ecotypes were designated Types A, B and C, and it was suggested that they might represent diVerent species. Type A, apparently the nominate form, lives in open water and preys mainly on Antarctic minke whales (Balaenoptera bonaerensis); Type B lives in loose pack ice where it preys mainly on seals (see also Visser et al. 2008), and Type C forages deep in the pack ice and among leads in the fast ice, and Wsh is its only known prey (see also Krahn et al. 2008).
123
1462
Little is known about either local or seasonal movement patterns of Antarctic killer whales. Current evidence suggests that Type A is probably migratory, moving to Antarctica during the southern summer to prey upon Antarctic minke whales, which also migrate there for the summer, then moving back to lower latitudes during the southern winter (Kasamatsu and Joyce 1995; Mikhalev et al. 1981). Less is known about the movements of the ice-inhabiting types (B and C). Both have been photographed once in New Zealand waters, for example, suggesting they could be migratory (Visser 1999; Pitman and Ensor 2003). Both forms normally forage in the ice, however, so either might be able to live in Antarctic waters year round, and in fact both have been documented there during the winter (Pitman and Ensor 2003). Types A and B killer whales are circumpolar in Antarctic waters, but Type C has only been reported from eastern Antarctica (Pitman and Ensor 2003). The Type C killer whale appears to be a summer resident in McMurdo Sound in the southern Ross Sea where it is by far the most common form in and around the icebreaker channel (see below). For example, one of us (RLP) saw one or more groups of Type C whales during 17 of 18 helicopter Xights out of McMurdo during the 2004–2005 and 2005–2006 seasons. Type B killer whales have been identiWed in the icebreaker channel only once, but they are sometimes seen in nearshore areas by penguin researchers based on Ross Island (Ballard and Ainley 2005; Ainley personal communication). During December–February each year, one or two icebreakers are present in McMurdo Sound to break a channel through the fast ice so supply ships can access McMurdo Station. Depending on the amount of ice cover in a given year, the distance through the fast ice between McMurdo Station and open water can range from 20 to over 150 km. Since at least the early 1970s, killer whales have been recorded annually in McMurdo Sound shortly after the icebreaking has begun. The whales have apparently learned to take advantage of foraging habitat made available when the icebreaker(s) opens up the channel, which allows whales to forage deeper into the fast ice than they could otherwise (Pitman and Ensor 2003). The occurrence of killer whales in the icebreaker channel and associated leads provides an unparalleled opportunity to approach within a few meters of free-swimming individuals and attach satellite transmitters. Until recently, the only killer whales that have been satellite tracked were individuals that were captured and instrumented on board a ship (Similä et al. 2001). During January–February 2006, we used an air-gun or crossbow to attach satellite tags to the dorsal Wn of Type C and Type B killer whales in McMurdo Sound. Here we present our tracking results and describe the movement patterns of Type B and Type C killer whales in the southern Ross Sea, and compare both to analogous transient and resident ecotypes in the eastern North PaciWc.
123
Polar Biol (2008) 31:1461–1468
Materials and methods Our operations were based out of McMurdo Station, on Ross Island, in the southern Ross Sea, Antarctica (Fig. 1a) from 14 January to 2 February 2006. We located killer whales by Xying in a helicopter along the icebreaker channel and leads opened by the icebreaker. When a group was located, we traveled approximately 3–5 km ahead in their direction of travel, landed the helicopter where the lead or channel narrowed, or near an area of open water in the channel, and waited for the whales to swim past us (Fig. 2). The satellite transmitters were surface-mounted with sub-dermal attachments (Andrews et al. 2005) and were based on the SPOT5 location-only Argos transmitter (Wildlife Computers, Redmond, WA). The transmitter electronics were cast in epoxy (Scotchcast 5, 3 M, Austin, TX) in three diVerent designs that we refer to based upon their shape and whether they were attached with one or two darts: (1) single-dart (tag dimensions: 4.8 cm £ 4.8 cm by 1.5 cm tall, 48 g); (2) low proWle double-dart (9.0 cm £ 3.2 cm by 1.5 cm tall, 56 g); (3) compact double-dart (6.0 cm £ 3.5 cm by 2.5 cm tall, 41 g; Fig. 3). All transmitters had an 18.0-cm antenna mounted on the side opposite from the attachment darts. The stainless steel barbed darts were designed to penetrate up to 6.5 cm into the dorsal Wn tissue (Fig. 3). The single-dart and compact double-dart transmitters were remotely attached to the dorsal Wn using an adjustable pressure, modiWed air-gun (Model JP.SP-25, Dan-Inject, Børkop, Denmark). The low proWle double-dart transmitter was delivered with a crossbow equipped with a 68.0 kg (150 lb) draw limb, with a 25.4 cm (10 in.) power stroke. The tag antenna was inserted into a hollow aluminum shaft or crossbow bolt. On contact with the dorsal Wn, the bolt or shaft fell away and was retrieved by a tether line, leaving the transmitter attached to the dorsal Wn. All transmitters were remotely attached by standing on the ice edge and launching the tag as a whale swam by within 1–5 m (Fig. 2). To conserve power, tag transmissions were limited by a submersion sensor to times when the whales were at the surface, but no more frequently than once per 30 s. The transmitters were also limited to only 200 transmissions per day and only during 00:00 hour–10:00 hours GMT. Transmissions were scheduled to occur daily during January and February, every Wfth day in March, April, and May, and every tenth day thereafter. With this duty-cycle the batteries should have provided enough power for 7–12 months of transmissions. Geographic locations of transmitters were determined by Service Argos (=CLS America) using the Doppler shift created by the satellite passing overhead. We determined the plausibility of each location using the Douglas Argos Wlter
Polar Biol (2008) 31:1461–1468
1463
Fig. 1 a The study area in the western Ross Sea, Antarctica, and satellite tracks of: b Type B killer whale # 57558; c Type C whale #57557; d Type C, whale #57556; e Type C whale # 47725; f Type C whale #57548, plotted with the margin of the land-fast ice edge on January 20 (dotted line) and February 15 (dashed line). Note that maps d–f are zoomed to a larger scale than the others
123
1464
Fig. 2 Attaching a satellite tag to a Type B killer whale in a fast ice lead in McMurdo Sound, southern Ross Sea, Antarctica in February 2006
Polar Biol (2008) 31:1461–1468
it is to immediately return back to the same vicinity with no intervening location Wxes on the outward or return path). Killer whales can travel at over 30 km h¡1 for short periods of up to 15 min (C. Matkin, personal communication), with an estimated maximum sustainable swimming speed of 20 km h¡1 (Guinet et al. 2007); therefore we chose 25 km h¡1 as the upper limit for our purposes. We calculated the 90% minimum convex polygon (MCP) area that was used or traversed by the whales using the Animal Movement Analysis ArcView Extension (Hooge and Eichenlaub 2000). This was done not to estimate a proper home range, but rather to provide a rough comparison of movement rates, area coverage and habitat use for the diVerent ecotypes. For the Type C whale tagged on 20 January 2006 [whale #57557] we used only the location points from the time of tagging until 06 March 2006, which was when that whale began a directed migration out of McMurdo Sound as the southern Ross Sea began to freeze up (Fig. 1c). Mean and maximum net daily travel distances were calculated for each whale using one location per day, the location with the best Argos location class for that day. The position of the ice edge (Fig. 1b–f) was determined from Ross Sea MODIS images courtesy of the MODIS Rapid Response Project at NASA/Goddard Space Flight Center.
Results
Fig. 3 The compact double-dart satellite transmitter design that resulted in the longest attachment durations
v.7.03 (Douglas 2007). This Wlter consists of a systematic algorithm that considers location class (LC), proximity to previous and subsequent locations, rate of movement, and the acuteness of the angle formed by the previous and consecutive locations. We considered locations to be plausible and retained them for analyses if the LC index was 3 or 2 or if the distance to the previous or subsequent location was
